U.S. patent number 11,346,795 [Application Number 16/949,041] was granted by the patent office on 2022-05-31 for xps metrology for process control in selective deposition.
This patent grant is currently assigned to NOVA MEASURING INSTRUMENTS, INC.. The grantee listed for this patent is NOVA MEASURING INSTRUMENTS, INC.. Invention is credited to Charles Thomas Larson, Wei Ti Lee, Kavita Shah.
United States Patent |
11,346,795 |
Larson , et al. |
May 31, 2022 |
XPS metrology for process control in selective deposition
Abstract
XPS spectra are used to analyze and monitor various steps in the
selective deposition process. A goodness of passivation value is
derived to analyze and quantify the quality of the passivation
step. A selectivity figure of merit value is derived to analyze and
quantify the selectivity of the deposition process, especially for
selective deposition in the presence of passivation. A ratio of the
selectivity figure of merit to maximum selectivity value can also
be used to characterize and monitor the deposition process.
Inventors: |
Larson; Charles Thomas
(Belmont, CA), Shah; Kavita (Mountain View, CA), Lee; Wei
Ti (San Jose, CA) |
Applicant: |
Name |
City |
State |
Country |
Type |
NOVA MEASURING INSTRUMENTS, INC. |
Santa Clara |
CA |
US |
|
|
Assignee: |
NOVA MEASURING INSTRUMENTS,
INC. (Fremont, CA)
|
Family
ID: |
1000006338986 |
Appl.
No.: |
16/949,041 |
Filed: |
October 9, 2020 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20210025839 A1 |
Jan 28, 2021 |
|
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
16351153 |
Mar 12, 2019 |
10801978 |
|
|
|
62641430 |
Mar 12, 2018 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01N
23/2273 (20130101); C23C 16/06 (20130101); C23C
16/24 (20130101); C23C 16/45529 (20130101); C23C
16/52 (20130101); H01L 21/67253 (20130101) |
Current International
Class: |
G01N
23/2273 (20180101); C23C 16/52 (20060101); H01L
21/67 (20060101); C23C 16/24 (20060101); C23C
16/06 (20060101); C23C 16/455 (20060101) |
Field of
Search: |
;250/305 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: McCormack; Jason L
Attorney, Agent or Firm: Womble Bond Dickinson (US) LLP
Bach, Esq.; Joseph
Parent Case Text
CROSS-REFERENCE TO RELATED APPLICATIONS
This application is a continuation of U.S. application Ser. No.
16/351,153, filed Mar. 12, 2019, which claims the benefit of U.S.
Provisional Application No. 62/641,430, filed Mar. 12, 2018, which
are hereby incorporated by reference in its entirety.
Claims
The invention claimed is:
1. A method for monitoring quality of process forming a second
layer having a second species over a first layer having first
species using X-ray photoelectron spectroscopy (XPS) measurements,
comprising: obtaining a first reference spectra of the first
species from an XPS measurement of a reference wafer having the
first layer; obtaining a second reference spectra of the first
species from an XPS measurement of the second layer deposited by a
prescribed number of deposition cycles on the first layer of the
reference wafer; generating a reference attenuation spectra
corresponding to attenuation of the second reference spectra of the
first species from the first reference spectra of the first
species; taking an XPS measurement of a sample wafer having the
first layer and generating first sample spectra of the first
species; taking an XPS measurement of the sample wafer after the
second layer has been deposited over the first layer by the
prescribed number of deposition cycles, and generating a second
sample spectra of the first species; generating a sample
attenuation spectra corresponding to attenuation of the second
sample spectra of the first species and the first sample spectra of
the first species; generating a quality of process value by
comparing the sample attenuation spectra and the reference
attenuation spectra.
2. The method of claim 1, wherein the first species comprises
silicon.
3. The method of claim 1, wherein the first layer comprises
SiO2.
4. The method of claim 1, wherein the first layer comprises a
dielectric pattern.
5. The method of claim 4, wherein the dielectric pattern comprises
SiO2.
6. The method of claim 1, wherein the first species and the second
species comprise metal.
7. The method of claim 1, wherein the second species comprise
hafnium.
8. The method of claim 1, wherein the second layer comprises
hafnium oxide.
9. The method of claim 1, wherein the second layer comprises
hafnium nitride.
10. The method of claim 1, wherein the second layer comprises metal
oxide.
11. The method of claim 1, wherein the first layer comprises SiO2
and the second layer comprises hafnium oxide.
12. A method utilizing X-ray photoelectron spectroscopy (XPS) for
monitoring quality of selective deposition process forming a second
layer having a second species over a first layer having first
species, comprising: a. prior to performing the selective
deposition process, taking an XPS measurement of a sample wafer
having the first layer and generating sample integrated intensity
of the first species; b. after the sample wafer undergone preset
cycles of selective deposition process, taking an XPS measurement
of the sample wafer and generating sample integrated intensity of
the second species; c. generating a sample value by taking a ratio
of the sample integrated intensity of the second species and the
sample integrated intensity of the first species; d. generating a
quality of process value by taking the ratio of the sample value
and a pre-generated reference value.
13. The method of claim 12, wherein step b is repeated multiple
times.
14. The method of claim 12, wherein the first species comprise
silicon.
15. The method of claim 14, wherein the first layer comprises
SiO2.
16. The method of claim 12, wherein the first layer comprises
patterned SiO2.
17. The method of claim 12, wherein the first and the second
species comprise metal.
18. The method of claim 12, wherein the second species comprise
hafnium.
19. The method of claim 12, wherein the second layer comprises at
least one of hafnium nitride, hafnium oxide, and metal oxide.
20. The method of claim 12, wherein the first layer comprises SiO2
and the second layer comprises hafnium oxide.
Description
BACKGROUND
1. Field
This disclosure relates generally to the field of process control
and monitoring in the semiconductor fabrication field. The
disclosed process control technique is particularly suitable for
monitoring selective deposition processes.
2. Related Art
For decades the semiconductor industry relied on photolithography
to generate the patterning required for the chips' circuitry.
Photolithography enabled depositing each layer over the entire
wafer, and then patterning the layer to form the circuitry. In
addition to adding many steps and cost to the chip fabrication
process, current nano-scale features make photolithography
incredibly difficult and, indeed, perhaps at some point impossible.
Additionally, double and multi-patterning used to define nano-scale
features (requiring two or more separate lithography and etch steps
to define a single layer) may lead to unacceptable edge placement
errors (EPE) and overlay misalignments.
An emerging technique, called Selective Deposition, deposits each
layer only at the areas of the designed circuitry, thus avoiding
the need for photolithography patterning. One promising example of
selective deposition is the use of Atomic Layer Deposition (ALD) to
repeatedly form Self-Assembly Monolayers (SAM), wherein each
monolayer is deposited only at the regions of the designed
circuitry. A similar technique, Molecular Layer Deposition (MLD) is
used for deposition of organic materials. Generally, the top
surface of the substrate has a dielectric pattern, a metal pattern,
and possibly a semiconductor pattern, and the next layer to be
formed may be a metal layer over the metal pattern, a dielectric
layer over the dielectric pattern, or a semiconductor over the
semiconductor pattern. This may require area activation or area
deactivation (passivation) prior to the next layer's formation. The
ALD deposition using SAM with surface passivation may be a
promising technique as it both avoids the photolithography step and
uses the surface's chemistry to make the alignments, thus
preventing EPE and overlay errors.
Regardless of which technique is used, metrology and process
control tools will be required in order to implement an integrated
process with acceptable yield. However, to date no suitable
metrology tools have been developed for process monitoring and
qualification. The conventional tools used in the labs today for
investigating these processes include Scanning Electron Microscopy
(SEM), Atomic Force Microscopy (AFM) and Tunneling Electron
Microscopy (TEM). These tools are too slow to be employed in a
production environment, and are incapable of providing real-time
monitoring of the process, so as to indicate a drift or a failure
of the process in a commercial fabrication setting.
X-ray photoelectron spectroscopy (XPS) has been used to analyze
surface chemistry of substrates. XPS spectra are obtained by
irradiating the substrate with a beam of X-rays, while
simultaneously measuring the kinetic energy and number of electrons
that escape from the top layers of the substrate. Similarly, X-ray
fluorescence (XRF) has been widely used for elemental and chemical
analysis of samples, by sampling the emission of characteristic
"secondary" (or fluorescent) X-rays from a material that has been
excited by bombarding with high-energy X-rays or gamma rays.
3. Problem to be Solved
In order to enable selective deposition in commercial fabrication
environment, a need exists in the art for process monitoring and
control. The methodology should provide a fast, direct,
non-destructive measurements of the quality of the process on the
wafer, to enable analysis of process quality and detection of
process drift.
SUMMARY
The following summary of the disclosure is included in order to
provide a basic understanding of some aspects and features of the
invention. This summary is not an extensive overview of the
invention and as such it is not intended to particularly identify
key or critical elements of the invention or to delineate the scope
of the invention. Its sole purpose is to present some concepts of
the invention in a simplified form as a prelude to the more
detailed description that is presented below.
Disclosed embodiments enable analyzing and monitoring deposition
and/or passivation processes, especially in the context of
selective deposition. The embodiments also provide figures of merit
to enable quantifying the quality of the process and identifying
process drifts or predicting required maintenance of the deposition
equipment. As such, the embodiments enable the implementation of
selective deposition in a commercial fabrication environment.
In the disclosed embodiments, XPS measurements are used to analyze
the quality of layers formed during selective deposition. The
measurements can be implemented during different steps of the
selective deposition, e.g., to analyze the quality of the
passivation, the quality of the deposited layer, the presence of
deposited material over the passivated areas, the presence of
pinholes, etc. As such, the disclosed embodiments help ensure that
material is deposited properly in areas where it should be
deposited, and that no material is deposited where it should not be
deposited.
Examples disclosed herein illustrate the use of relative XPS
intensities to determine metrics for process results, such as
passivation and degree of selectivity.
In the context of selective deposition process, disclosed
embodiments enable quantification of the passivation and the
selectivity. In one aspect, the passivation is evaluated by
generating a metric referred to as goodness of passivation (GoP),
which enables evaluation of the passivation of a sample against
optimal or required passivation quality. The GoP is a functional
operation of XPS intensities. For example, the GoP may be a "ratio
of ratios", by setting the functional operation to be a ratio of
two integrated intensities ratios. In such an example, the ratio
may be a linear ratio, a ratio of the squares of the XPS
intensities, a ratio of the square root of the XPS intensities,
etc. Thus, the use herein of the term ratio is intended to
encompass any of these functional ratios.
In one example, first, a reference ratio is generated by performing
XPS measurements of a reference wafer, and taking the ratio of the
integrated intensity of the passivation atomic species over the
passivated atomic species. For example, when the passivated layer
is metal lines, e.g., copper lines, and the passivation is
undecanethiol (UDT), the passivation atomic species would be sulfur
and the passivated atomic species would be copper. During the
selective deposition process, after passivation has been formed on
a sample wafer, XPS measurements are taken of the sample wafer and
a sample ratio is generated by taking the ratio of the integrated
intensity of the passivation atomic species over the passivated
species for the sample wafer. The GoP is generated by a functional
operation, e.g., taking the ratio, of the sample ratio and the
reference ratio.
The reference wafer may incorporate both layers, such that the
measurements may be taken concurrently. For the example of a
Sulphur based passivation of copper, the reference sample may be a
silicon wafer with a blanket Cu layer (or large regions of Cu, such
as metrology pads) and an S based passivation layer on the top of
the Cu. The XPS reference data from the Cu and S would be collected
at the same time (or nearly at the same time).
According to further aspects, the selectivity of the deposition
process is quantified. For example, when performing atomic layer
deposition (ALD) of a second dielectric over an underlying low-k
dielectric, in the presence of passivated metal lines, it is
desirable to verify that the second dielectric is indeed deposited
over the low-k dielectric, and that no dielectric is deposited over
the passivated metal lines. To that effect, disclosed embodiments
provide a Selectivity Figure of Merit (SFM) that can be used to
determine whether an acceptable selectivity has been achieved. The
SFM is a functional operation, e.g., a ratio of two ratios. First,
XPS measurements of a reference wafer are taken and are used to
derive two reference values: reference integrated intensity of the
metal lines species (e.g., copper) and reference integrated
intensity of the underlying dielectric species (e.g., silicon).
After an x cycles of ALD selective deposition process are performed
on a sample wafer, XPS measurements of the sample wafer are taken
and are used to derive two sample values: sample integrated
intensity of the metal lines species (e.g., copper) and sample
integrated intensity of the underlying dielectric species (e.g.,
silicon). A first species ratio is obtained by taking the ratio of
the sample integrated intensity of the metal lines species over the
reference integrated intensity of the metal lines species. For
perfect selectivity this ratio should approach 1, indicating that
no dielectric has been deposited over the metal lines. A second
species ratio is obtained by taking the ratio of the sample
integrated intensity of the dielectric species over the reference
integrated intensity of the dielectric species. The integrated
intensity of the dielectric species would be attenuated with
increased number of ALD cycles. The SFM is obtained by a functional
operation, e.g., taking the ratio of the first species ratio over
the second species ratio.
BRIEF DESCRIPTION OF THE DRAWINGS
Other aspects and features of the invention would be apparent from
the detailed description, which is made with reference to the
following drawings. It should be appreciated that the detailed
description and the drawings provides various non-limiting examples
of various embodiments of the invention, which is defined by the
appended claims.
The accompanying drawings, which are incorporated in and constitute
a part of this specification, exemplify the embodiments of the
present invention and, together with the description, serve to
explain and illustrate principles of the invention. The drawings
are intended to illustrate major features of the exemplary
embodiments in a diagrammatic manner. The drawings are not intended
to depict every feature of actual embodiments nor relative
dimensions of the depicted elements, and are not drawn to
scale.
FIG. 1 is a partial cross-section of a prior art structure for
selective deposition.
FIG. 2 illustrates a selective deposition structure according to a
disclosed embodiment.
FIG. 3 illustrates XPS spectra indicating peaks corresponding to
four species, according to disclosed embodiment.
FIGS. 4A-4C illustrate wafer mapping of GoP according to disclosed
embodiment.
FIG. 5 illustrates an example of an apparatus according to one
embodiment.
DETAILED DESCRIPTION
Embodiments of the inventive deposition process control and
monitoring will now be described with reference to the drawings.
Different embodiments or their combinations may be used for
different applications or to achieve different benefits. Depending
on the outcome sought to be achieved, different features disclosed
herein may be utilized partially or to their fullest, alone or in
combination with other features, balancing advantages with
requirements and constraints. Therefore, certain benefits will be
highlighted with reference to different embodiments, but are not
limited to the disclosed embodiments. That is, the features
disclosed herein are not limited to the embodiment within which
they are described, but may be "mixed and matched" with other
features and incorporated in other embodiments.
FIG. 1 illustrates a cross-section of a prior art device
demonstrating a selective deposition process. The substrate 100 may
include other layers between the top surface and the illustrated
circuitry on top of the substrate, but these layers are not
relevant to the subject discussion and are, therefore, not shown.
The circuitry shown in the example of FIG. 1 includes metal traces
105, a passivation layer 110 over the metal traces 105, first
dielectric layer 115, and a second dielectric layer 120. An example
of a process for fabricating the structure shown in FIG. 1 includes
the deposition and patterning of the metal traces 105, e.g., metal
lines inside trenches formed in the first dielectric layer 115.
Then, the metal traces 105 are passivated so that when the second
dielectric layer 120 is formed by selective deposition, no
dielectric material will attach to the metal lines. After the
formation of passivation layer 110, the second dielectric layer 120
is formed by, e.g., ALD selective deposition. After forming the
second dielectric layer 120, the passivation layer 110 can be
removed. However, during this process there are at least two steps
during which the process can fail. First, if the passivation layer
is not properly formed, dielectric material may form over the metal
lines. Second, if the second dielectric layer is insufficiently
thick, or has voids, it may lead to device failure. Thus, it is
desirable to be able to non-destructively test the wafer during
fabrication to avoid such failures.
According to some embodiments, SAM of undecanethiol (UDT) is used
as passivation over the metal traces to enable selective deposition
of second dielectric layer over an underlying dielectric layer. The
UDT may be, e.g., 11-amino-1-undecanethiol, 11-mercapto-1-undecanol
(OH--C11), and 11-amino-1-undecanethiol hydrochloride (NH2-C11). By
forming several layers of SAM UDT, it prevents the dielectric from
attaching to the metal layer, thus selectively depositing
dielectric only over the underline dielectric areas. However,
according to disclosed embodiments, it is preferable to determine
the quality of the SAM UDT prior to performing the selective
deposition of the dielectric, so as to ensure no dielectric adheres
to the metal.
In a first example, XPS is used to determine the quality of a
deposited layer over a blanket wafer, which can be used, e.g., to
generate a baseline for future testing. In this example, the wafer
is coated with a uniform layer of copper, and ALD is used to form a
layer of HfNx (Hafnium nitride) over the entire copper layer.
First, XPS is used to obtain the signal spectrum obtained from the
wafer with only the copper layer. Then the spectrum is obtained
after an intermediate number of ALD cycles, say ten cycles of HfNx
deposition. The intensity of the spectra for the copper signal is
expected to drop following successive deposition, as the electrons
from the copper layer now must travel through a certain thickness
of HfNx layer. At the same time, the intensity of the HfNx specie
should increase, indicating the increased amount of HfNx present
over the copper. This is repeated over certain number of cycles of
ALD, such that as the intensity of the spectra from the copper is
diminished, the spectra from the HfNx layer intensifies. Thus, one
obtains a baseline of the signals for an HfNx layer growing over
copper. By comparing spectra taken during future manufacturing
cycles with these spectra, one can monitor the formation of the
deposited layer.
In a separate process, the copper is first passivated with SAM UDT
prior to performing the ALD HfNx growth and measurements cycles. In
this embodiment the top layer of the substrate includes lines of
copper interspersed with areas of dielectric material. In a first
step, the copper lines are passivated by forming several layers of
SAM UDT. One or more XPS spectra may be taken to obtain the
signature of the copper emission through the UDT layers.
Thereafter, selective deposition of dielectric HfNx is performed
and one or more spectra may be taken to obtain signature of the
copper emission and the dielectric emission. The copper emission
may be used as calibration for future investigation whether
dielectric is being deposited over the copper or determine
degradation of the UDT, while the dielectric emission can be used
as calibration for future examination of the quality of the
deposited dielectric HfNx over the dielectric.
The discussion will now refer to FIG. 2 to present an embodiment
that provides a figure of merit for the quality of a passivation
layer. The same process can be implemented to provide a figure of
merit for a selective deposition process, but as the passivation
process may prove essential in selective deposition technology, a
first example is provided with respect to passivation, and is
referred to a "goodness of passivation" (GoP) measurement. The GoP
value can be used to monitor the process for the quality of the
passivation and for undesirable dielectric deposition over the
passivation. The passivation may be, e.g., a SAM UDT or other
organic or inorganic layers. Notably, in case of UDT, which is
formed of long chains of polymeric molecules, the formed layers are
optically transparent, such that optical inspection of the DUT
layers is not possible.
In FIG. 2, metal lines 205, e.g., copper, are interspersed between
first dielectric areas 215, e.g., low-k SiO2, etc., over substrate
200. Passivation 210 may be, e.g., SAM UDT, and the selectively
deposited second dielectric 220 may be, e.g., HfNx. Some circuitry
layers may exist between the substrate 200 and the first dielectric
215, but is omitted for clarity. As illustrated in FIG. 2, while
ideally the UDT would passivate only the metal lines 205 and the
HfNx would deposit only on top of the first dielectric 215, in this
case some of the second dielectric 220 has been deposited over the
passivation 210, and is identified as 220'. As the deposition of
dielectric 220' is undesirable, an objective of this embodiment is
to identify how much of dielectric 220' is present on the
passivation 210, i.e., over the metal lines 205.
As illustrated in FIG. 2, when taking XPS spectra, some
photoelectrons emitted from the metal line 205, identified as
e.sup.-.sub.1, travel only through passivation layer 210, and thus
are attenuated accordingly. Conversely, some photoelectrons,
identified as e.sup.-.sub.2 travel through both the passivation
layer 210 and the dielectric 220', and are attenuated to a
different degree.
FIG. 3 illustrates hypothetical spectra taken from a structure such
as that shown in FIG. 2. The abscissa indicates the binding energy
in electron volts, while the ordinate indicates the intensity in
counts per second. Thus, the signal along the abscissa indicates
what atomic species are present in the sample, while the intensity
indicates the abundance of that specie.
In one series of experiments, wafers with blanket copper layer were
used to deposit a blanket HfNx layer, with and without a layer of
UDT. XPS intensities for Cu3p were taken at different number of ALD
HfNx deposition cycles. Of course, with UDT over the copper no HfNx
deposition is expected, hence no attenuation of the Cu3p intensity
is expected. The experiments were repeated using patterned wafers
with copper lines over low-k dielectric, with and without UDT over
the copper lines. For blanket wafers without UDT, as the number of
HfNx ALD cycles increased, the Cu3p signal decreased while the Hf4f
signal increased, indicating HfNx deposition over the copper. Also,
there was drastic attenuation of the Cu3p signal for HfNx ALD
cycles on patterned wafer when no UDT was present, thereby
indicating that the HfNx was deposited on the copper when no UDT
was present on the copper lines. On the other hand, it was noted
that there was little Cu3p signal attenuation after up to 80 ALD
cycles of HfNx when UDT was present over the copper, for both
blanket and patterned wafers. This confirmed that UDT acts to
prevent HfNx deposition over the copper.
However, it was unexpectedly discovered that when taking the XPS
spectra after UDT, but prior to performing the HfNx ALD, the
attenuation of the Cu3p signal was much more drastic for blanket
wafers covered with UDT than for patterned wafer covered with UDT,
indicating that the UDT coverage of the copper lines of the
patterned wafer was not as good as the coverage of blanket wafers.
In fact, while the intensity of the Cu3p dropped by an order of
magnitude on blanket wafers after forming the UDT passivation,
there was much smaller intensity attenuation for the corresponding
patterned wafers. The inventors then postulated, and have then
unexpectedly discovered, that using the ratios of the XPS
intensities, the quality of the selective deposition process,
including the passivation, can be evaluated and monitored.
Similar experiments were made with blanket wafers having a layer of
low-k dielectric covering the entire wafer. XPS spectra were taken
over several ALD cycles of HfNx deposition, with and without an
intervening UDT layer. It was noted that the Si2p signal from the
underlying low-k dielectric was attenuating as the number of ALD
cycles increased, with or without the intervening UDT. It was also
noted that the Hf4f signal was increasing with or without the
intervening UDT. It was thus confirmed that the UDT does not
passivate the low-k dielectric.
As can be understood, the quality of the passivation is critical to
the selective deposition process. Accordingly, the inventors have
developed a method for characterizing the quality of the
passivation, which is referred to herein as Goodness of Passivation
(GoP). The GoP is a function of the XPS signal intensities measured
at different stages of the process. In this example, the function
is a ratio (linear, quadratic, etc.) of XPS signal intensities and
is a unitless fraction between zero and one, i.e.,
0.ltoreq.GoP.ltoreq.1. The GoP can be expressed as
GoP=f{(I.sub.P/I.sub.Subst).sub.s,(I.sub.P/I.sub.Subst).sub.0},
which, using the linear ratio function, is expressed as
GoP=(I.sub.P/I.sub.Subst).sub.s/(I.sub.P/I.sub.Subst).sub.0;
wherein I.sub.P is the XPS signal intensity of a specie forming the
passivation (I being the integrated intensity of the XPS spectrum
for a selected specie), I.sub.Subst is the XPS signal intensity of
a specie from the underlying layer, e.g., copper,
(I.sub.P/I.sub.Subst).sub.s is the measurements of the sample of
interest, and (I.sub.P/I.sub.Subst).sub.0 is the measurements of a
reference perfect passivation. Thus, if the XPS measurements of the
sample result in the same values as those of the reference, the GoP
equals 1. On the other hand, as the measurements of the sample
relate to inferior passivation, the GoP results in a number that is
smaller than 1, and is decreasing with the inferiority of the
passivation.
Applying the method of GoP to the embodiment illustrated in FIG. 2,
the spectra of sulfur is selected as the species of the passivation
layer and the spectra of copper is selected as the specie for the
underlying layer. Thus, the Goodness of Passivation relationship
becomes
GoP=(I.sub.Passivation/I.sub.Metal).sub.s/(I.sub.Passivation/I.sub.Metal)-
.sub.0=(I.sub.S/I.sub.Cu).sub.s/(I.sub.S/I.sub.Cu).sub.0; such that
with improved passivation I.sub.Passivation increases and
I.sub.Metal decreases, and when the measured passivation reaches
maximum quality the GoP ratio becomes 1, i.e.,
(I.sub.Passivation/I.sub.Metal).sub.s=(I.sub.Passivation/I.sub.Metal).sub-
.0. The values for (I.sub.Passivation/I.sub.Metal).sub.0 can be
obtained by measurements of passivation on a blanket copper
wafer.
As one example, the following embodiment utilizes the GoP
relationship to monitor the process in-line. Specifically, XPS
spectra of a substrate having a known perfect passivation are
obtained, to provide values for (I.sub.P/I.sub.Subst).sub.0. An
acceptable ratio threshold is set for GoP, under which it is
believed that the process results in unacceptable quality of
passivation. The process proceeds by taking XPS spectra of wafers
prior to and after passivation, to calculate GoP.sub.i for each or
every x number of wafers being processed. The resulting GoP.sub.i
is compared to the threshold. When GoP.sub.i falls below the
threshold, an alarm is issued and the passivation process may be
halted to investigate the reasons for the resulting low quality
passivation.
After the quality of passivation has been confirmed by an
acceptable GoP value, the wafers may proceed to the selective
deposition step. Generally, it is considered that the second
dielectric would indeed be deposited selectively only over the
areas of the first dielectric. Nevertheless, should this step be
required to be monitored, the GoP measurement can be used. That is,
if during the ALD process the second dielectric material is being
deposited on the passivation layer, it will change the value of the
GoP, thus indicating problem in the process. To illustrate, the GoP
measurement can be done on the wafer prior to the start of the
deposition as a reference value, and then during deposition to
check whether the value changes during or after the deposition
process. If ALD process is depositing HfO on the passivation layer,
the GoP will likely change as the HfO will attenuate the Cu and S
signals differently.
Alternatively, or additionally, a measure can be derived as
Goodness of Selective Deposition (GoSD), which is a function of XPS
intensities of the passivation and deposition species. The GoSD may
be a ratio (linear, squared, etc.), wherein in a linear fraction
form it can be expressed as
GoSD=(I.sub.SD/I.sub.P).sub.s/(I.sub.SD/I.sub.P); wherein I.sub.P
is the XPS signal intensity of the passivation specie and I.sub.SD
is the XPS signal intensity of the selective deposition specie. The
reference values are (I.sub.SD/I.sub.P).sub.0, while the sample
values are (I.sub.SD/I.sub.P).sub.s.
The GoSD can also be determined using XPS of the deposition and
underlying species. To illustrate, in the example of ALD deposition
of HfO, first HfO is deposited on a reference wafer having blanket
layer or large pads of SiO2. An XPS measurement is performed on the
reference wafer to obtain (I.sub.Hf/I.sub.Si).sub.0. Then a
patterned wafer is passivated and HfO2 is deposited over the
passivated wafer. An XPS measurement is taken after deposition to
obtain the values (I.sub.Hf/I.sub.Si).sub.s. The two ratios should
be the same if the deposition was selective. If the ratio resulting
from the patterned wafer is higher, it indicates that HfO grew on
the metal, meaning failed passivation. If the value is lower, it
means incomplete deposition of HfO2 over the SiO2.
In the selective deposition process integration scheme, once the
second dielectric selective deposition step is completed, the
passivation layer is removed. In that step, the GoP measurement can
be used to monitor how well the passivation was removed, except
that the intensity measurements that are used are from XPS signals
taken before and after removal of the passivation. Thus, the
relationship may be expressed as goodness of passivation
removal--GoPR=(I.sub.PR/I.sub.SD).sub.s/(I.sub.PR/I.sub.SD).sub.0;
wherein I.sub.PR is the XPS signal intensity after passivation
removal and I.sub.SD is the XPS signal intensity after the
selective deposition process.
Ultimately, what is required for a successful selective deposition
process is high selectivity, meaning proper deposition in the areas
where the deposition is required to occur, and no deposition on
areas where no deposition is required to occur. When the
selectivity is between a first material having element A and a
second material having element B, the inventors have developed a
Selectivity Figure of Merit (SFM) which is a function of XPS
intensities of elements prior to the start and after a number of
cycles of deposition. In one example the SFM is implemented using
ratios of XPS signal intensities, as follows:
SFM=(I.sub.Ai/I.sub.A0)/(I.sub.Bi/I.sub.B0), wherein I.sub.Ai is
the signal intensity from element A after i cycles of selective
deposition, I.sub.A0 is the signal intensity from element A prior
to start of selective deposition, I.sub.Bi is the signal intensity
from element B after i cycles of selective deposition, and I.sub.B0
is the signal intensity from element B prior to start of selective
deposition.
For example, applying the SFM formula to the embodiment of FIG. 2,
wherein metal copper lines are interspersed on a low-k first
dielectric, a SAM UDT passivation is formed over the copper lines,
and x cycles of selective deposition of HfNx, one gets:
SFMx=(I.sub.Cux/I.sub.Cu0)/(I.sub.Six/I.sub.Si0); wherein I.sub.Cux
is the signal intensity of Cu3p after x cycles of selective
deposition, I.sub.Cu0 is the signal intensity of Cu3p prior to
start of selective deposition (but after passivation), I.sub.Six is
the signal intensity from Si2p after x cycles of selective
deposition, and I.sub.Si0 is the signal intensity from Si2p prior
to start of selective deposition. That is, the values I.sub.Cu0 and
I.sub.Si0 are obtained from a passivated wafer prior to the start
of the deposition. The values I.sub.Cux and I.sub.Six are obtained
from the same wafer, after x cycles of selective deposition.
When the HfNx is deposited only over the low-k dielectric, i.e.,
perfect selectivity, the ratio I.sub.Cux/I.sub.Cu0 should equal or
approach 1, i.e., no attenuation of the XPS signal from copper.
Under such conditions the intensity attenuation of Si would depend
on ideal HfNx thickness at x cycles of selective deposition.
Therefore, the SFM correlates to both avoidance of deposition of
HfNx over the copper and the quality of HfNx deposition over the
low-k dielectric.
The data obtained from the experiments on blanket wafers can be
used to compare to the data obtained from testing on patterned
wafers. For example, the intensities obtained for Cu3p when forming
HfNx deposition over blanket copper wafer can be used to generate
ideal value for I.sub.Cux/I.sub.Cu0. Similarly, intensity values
obtained for Si2p when forming HfNx deposition over blanket low-k
wafer can be used to obtain ideal value for I.sub.Six/I.sub.Si0.
The "perfect" or maximum selectivity can be calculated by setting
I.sub.Cux/I.sub.Cu0=1 and using the blanket wafer data to set the
value for I.sub.Six/I.sub.Si0, resulting in
SFM.sub.MAX=1/(I.sub.Six/I.sub.Si0).sub.MAX, i.e., SFM.sub.MAX
equals the reciprocal (or multiplicative inverse) of
(I.sub.Six/I.sub.Si0).sub.MAX. The resulting SFM.sub.MAX value is
the best selectivity to be achieved in the process, and one may use
that value to derive a threshold value beyond which the process is
said to be out of spec and the process halted and investigated for
unacceptable drift. Additionally, the ratio of
SFM.sub.i/SFM.sub.MAX can be used to monitor the process, wherein
SFM is the SFM obtained for the i.sup.th measured wafer and
SFM.sub.MAX is the maximum selectivity value obtained from blanket
wafers. A threshold value can be derived for the ratio of
SFM.sub.i/SFM.sub.MAX beyond which the process is said to be out of
bounds.
As can be understood from the above, an in-line inspection and
monitoring method has been developed, using XPS spectra taken
during various steps of the selective deposition process. The
method quantifies the quality of the passivation by providing a
Goodness of Passivation (GoP) measurement. The method also
quantifies the deposition selectivity by providing a Selectivity
Figure of Merit (SFM) and a ratio of the SFM over the maximum
achievable selectivity. By setting the allowable values for the
GoP, SFM and SFM/SFM.sub.Max, the process can be monitored in-line
and excursions from the allowable values can be flagged to enable
investigation into the cause of the excursion. The allowable values
for the GoP, SFM and SFM/SFM.sub.Max, can be stored in the memory
of the XPS system, such that the XPS system can issue an alarm when
it detects excursion. Conversely, the allowable values for the GoP,
SFM and SFM/SFM.sub.Max, can be stored on a stand-alone computer,
together with the measurement program, such that it can obtain the
data from the XPS system and perform the analysis to identify
excursions of the process.
Moreover, the XPS measurements can be done at multiple sites across
the wafer and used to generate a map of GoP, SFM and
SFM/SFM.sub.Max across the wafer to identify non-uniformities in
the process. FIGS. 4A-4C illustrate an example wherein FIG. 4B
identifies enumerated positions of XPS measurements, FIG. 4A
illustrate a plot of the GoP value obtained for each position, and
FIG. 4C illustrate a map of the GoP values at each location of the
wafer. Similar plots and maps can be generated for the SFM and
SFM/SFM.sub.Max values.
FIG. 5 illustrates an example of an XPS apparatus according to one
embodiment. A photon source 505 illuminates a spot on wafer 500.
Photoelectrons are emitted from the wafer 500 and enter an electron
kinetic energy analyzer 510. Electrons within the selected energy
level exit the analyzer 510 and are collected by the electron
detector 515. The signal from the detector 515 is fed to the
computer 520, which calculates the GoP, SFM and SFM/SFM.sub.MAX.
The computer 520 may issue alarm upon detecting a process
excursion, and/or generate maps as shown in FIGS. 4A-4C.
As shown in FIG. 5, the computer 520 may be a part of the XPS
system, may be a part of a manufacturing control system, or may be
a stand along computer programmed to execute the embodiments
disclosed herein. As such, the invention includes a
computer-implemented method for evaluating process integrity in a
selective deposition processing environment forming a second layer
having a second atomic specie over a first layer having a first
atomic specie, comprising executing on a processor the steps of:
receiving reference XPS spectra of a reference wafer and generating
from the XPS spectra a reference integrated intensity of the first
specie and a reference integrated intensity of the second specie;
generating a reference value by taking the ratio of the reference
integrated intensity of the second specie and the reference
integrated intensity of the first specie; receiving sample XPS
spectra of a sample wafer and generating from the XPS spectra a
sample integrated intensity of the first specie and a sample
integrated intensity of the second specie; generating a sample
value by taking the ratio of the sample integrated intensity of the
second specie and the sample integrated intensity of the first
specie; and, generating a quality of process value by taking the
ratio of the sample value and the reference value.
In terms of process integration, an embodiment is provided wherein
the XPS spectra and analysis is incorporated as an in-line
monitoring method to quantify the quality of the process and
identify excursions and/or drifts of the process. The embodiment
starts with a wafer having a dielectric top layer and metal lines
interspersed in the dielectric layer. In one example the dielectric
top layer is a low-k dielectric and the metal lines comprise
patterned copper lines. The dielectric top layer is referred to as
dielectric underlayer, as the objective of the process is to form a
second dielectric layer over the dielectric top layer.
The embodiment proceeds to perform metal passivation on the wafer,
e.g., placing the wafer in a fabrication chamber to form SAM UDT
over the patterned copper lines. When the formation of the
passivation is completed, the wafer is transported to an XPS system
and XPS spectrum is obtained from the wafer. A sample ratio is then
obtained by obtaining an integrated intensity of the signal from
the passivation specie and dividing by the integrated intensity of
the signal from the metal specie. In the illustrated example the
sample ratio is obtained by taking the integrated intensity of
sulfur and dividing it by the integrated intensity of copper. A
goodness of passivation value is then obtained by taking the sample
ratio and dividing it by a reference ratio, wherein the reference
ratio is obtained by taking an integrated intensity of the signal
from the passivation specie and dividing by the integrated
intensity of the signal from the metal specie of a properly
passivated reference wafer. The goodness of passivation value is
then used to characterize the quality of the passivation by, e.g.,
comparing it to a preset threshold, comparing it to historical
values, etc. Also, the XPS system can be used to take several
spectra at multiple locations over the wafer, such that multiple
goodness of passivation values can be calculated, one for each XPS
spectrum location. Then a table or a map of the values can be
provided and used to analyze the uniformity of the passivation over
the wafer and the quality of the passivation process wafer to
wafer.
When the goodness of passivation value(s) are acceptable, in a
further embodiment the process proceeds to forming a second
dielectric layer, e.g., by transporting the wafer into an ALD
system and performing ALD selective deposition of dielectric
material, e.g., HfNx. When the second dielectric layer has been
formed, the wafer is transported again to an XPS system and at
least one XPS spectrum is taken. From the XPS spectrum, an
integrated intensity of metal specie is calculated, and an
integrated intensity of the underlying dielectric specie is
calculated. In the illustrated example the metal specie is copper
and the underlying dielectric specie is silicon. A metal intensity
ratio is then calculated by dividing the integrated intensity of
metal specie by a reference integrated intensity of metal specie
that was obtained from a reference wafer. A dielectric intensity
ratio is similarly calculated by dividing the integrated intensity
of the dielectric specie by a reference integrated intensity of
dielectric specie that was obtained from the reference wafer. A
selectivity figure of merit is calculated by dividing the metal
intensity ratio by the dielectric intensity ratio, and is used to
characterize the quality of the selective deposition process by,
e.g., comparing it to a preset threshold, comparing it to
historical values, etc. Also, the XPS system can be used to take
several spectra at multiple locations over the wafer, such that
multiple selectivity figure of merit values can be calculated, one
for each XPS spectrum location. Then a table or a map of the values
can be provided and used to analyze the uniformity of the
dielectric deposition over the wafer and the quality of the
selective deposition process wafer to wafer.
In a further embodiment, an XPS system is provided, comprising: an
x-ray source configured to illuminate a spot on a wafer; an
electron kinetic energy analyzer configured to separate electrons
emitted from the spot according to kinetic energy; an electron
detector detecting electrons exiting the electron kinetic energy
analyzer and generating detection signal; a controller receiving
the detection signal and comprising a memory storage and a
processor, the memory storage having a reference ratio stored
therein; the memory storage further having a program stored therein
which, when executed, causes the processor to perform the
operations comprising: operating on the detection signal to
generate integrated intensity of a first specie and generating
integrated intensity of a second specie; generate sample ratio by
dividing the integrated intensity of the first specie by the
integrated intensity of the second specie; and generating a quality
value by dividing the sample ratio by the reference ratio.
In a further embodiment, the memory storage has further stored
therein a reference second specie intensity and a reference third
specie intensity, and the program further causes the processor to
perform the operations comprising: operating on the detection
signal to generate integrated intensity of a third specie; generate
a second species ratio by dividing the integrated intensity of the
second specie by the reference second species intensity; generate a
third species ratio by dividing the integrated intensity of the
third specie by the reference third species intensity; generate a
selectivity value by dividing the second specie ratio by the third
specie ratio.
The invention also includes a computer-implemented method for
evaluating process selectivity in a selective deposition processing
environment forming a second layer over a first layer, wherein the
first layer comprises a pattern of a first material having a first
atomic specie and a second material having a second atomic specie,
comprising executing on a processor the steps of: receiving
reference XPS spectra of a reference wafer and generating from the
XPS spectra a reference integrated intensity of the first specie
and a reference integrated intensity of the second specie;
receiving sample XPS spectra of a sample wafer and generating from
the XPS spectra a sample integrated intensity of the first specie
and a sample integrated intensity of the second specie; generating
a first species value by taking a ratio of the sample integrated
intensity of the first specie and the reference integrated
intensity of the first specie; generating a second species value by
taking a ratio of the sample integrated intensity of the second
specie and the reference integrated intensity of the second specie;
generating a selectivity value by taking the ratio of the first
species value and the second species value; and, generating a
quality of process value by taking the ratio of the sample value
and the reference value.
In the context of this disclosure, a low-k dielectric is a material
with a smaller dielectric constant than silicon dioxide. Low-k
dielectric can be formed by doping silicon dioxide with fluorine or
carbon, by generating pores in silicon dioxide, or by using spin-on
silicon based polymeric dielectric materials. The resulting low-k
dielectric has dielectric constant lower than 3.9.
It should be understood that processes and techniques described
herein are not inherently related to any particular apparatus and
may be implemented by any suitable combination of components.
Further, various types of general purpose devices may be used in
accordance with the teachings described herein. The present
invention has been described in relation to particular examples,
which are intended in all respects to be illustrative rather than
restrictive. Those skilled in the art will appreciate that many
different combinations will be suitable for practicing the present
invention.
Moreover, other implementations of the invention will be apparent
to those skilled in the art from consideration of the specification
and practice of the invention disclosed herein. Various aspects
and/or components of the described embodiments may be used singly
or in any combination. It is intended that the specification and
examples be considered as exemplary only, with a true scope and
spirit of the invention being indicated by the following
claims.
* * * * *